Published Articles

Other Publications

Working Papers and Manuscripts

Comments are greatly appreciated.

  • Reeves, Andrew, Wendy K. Tam Cho, and James G. Gimpel. "Ecologies of Unease: Geographic Context and National Economic Evaluations"

    Abstract

    Assessment of the nation?s economic performance has been repeatedly linked to voters? decision making in presidential elections. Here we inquire as to where those economic evaluations originate. One possibility in the politicized environment of a major campaign is that they are partisan determinations and do not directly reflect actual economic circumstances. Another possibility is that these judgments arise from close attention to news media, which is presumably highlighting national economic conditions as a facet of campaign coverage. Still a third explanation is that voters derive their national economic evaluations from living out their lives in particular localities which may or may not be experiencing the conditions that affect the nation as a whole. As a container of relevant experience, locality may vary from the official institutional boundaries that are commonly assumed to function as such. Drawing upon data from the 2008 presidential election, we find that varying local conditions do shape the economic evaluations of political independents, with changing gas prices figuring more prominently in those evaluations than either accumulated foreclosures or rising unemployment. Local economic experience does appear to shape national economic evaluations for those who aren?t making up their minds on partisan grounds.

  • Gasper, John, and Andrew Reeves. "Make it Rain?: Governors, Presidents, and the Electoral Consequences of Severe Weather"

    Abstract

    Democratic accountability requires that voters make reasonable evaluations and judgements of incumbent performance. Are incumbent politicians blamed for occur- rences that are well beyond human control, or are voters more sophisticated taking context and the e?orts of the incumbent into account? We ?nd evidence of voter re- sponsiveness to how presidents and governors deal with the fall out from disasters; however, we also ?nd that voters are sensitive to the very occurrence of severe weather damage. In a county level analysis of election returns from 1970 to 2006, we show that governors and presidents are rewarded for disaster declarations to the tune of ap- proximately two and half and six-tenths of a percentage point respectively. When the governor request a disaster declaration and the president denies the request, governors are rewarded with almost three percentage points while presidents are punished by just under half a point. Our analysis shows that in most cases, the cost of blame for disaster damage far exceeds the electoral gain from a declaration.
  • Ansolabehere, Stephen, Andrew Reeves, and Stephen Graves. "Probable Winners: A Probability Model and Empirical Analysis of Reversals and Accuracy in Election Recounts"

    Abstract

    This paper examines the probability that recounts reverse initial election outcomes. We develop a counting probability model of election reversals from recounts, which leads to a normal approximation for assessing the likelihood that a recount reverses an initial count. We confirm the model by testing two basic empirical implications of the model using data from New Hampshire towns. We further show how the model may be adapted to more specific circumstances, including variations in the mode of counting and possible asymmetries between candidates in the rate of recovery of votes. Based on an analysis of recounted ballots from 1946 through 2004 in the state of New Hampshire, we find that the discrepancies between initial counts and recounts is nearly twice that for hand counted ballots as for machine counted ballots (1.55 compared to .78). We also find that election recounts rarely change the winners.

  • Reeves, Andrew. "Political Disaster? Electoral Politics and Presidential Disaster Declarations" (Supplemental materials)

    Abstract

    Presidential disaster declarations allow the president to unilaterally clear the way for potentially billions of dollars to be distributed to specific constituencies. In an analysis extending from 1981 to 2004, I find that electoral factors (as measured by electoral vote size and statewide competitiveness) influence whether a president issues a disaster declaration. For example, a highly competitive state can expect to receive over 60% more presidential disaster declarations than an uncompetitive state, holding all else constant including the damage caused by the disaster. The relationship between competitiveness and disaster declarations has only existed since the late 1980s, before which time there was actually a negative (although statistically insignificant) relationship. Additionally, I find that these decisions have the intended electoral benefits --- voters react and reward presidents for presidential disaster declarations. A president can expect a 1.7% increase in a statewide contest in return for a single presidential disaster declaration.

    Software

  • muRL. Provides mailmerge methods for reading spreadsheets of addresses and other relevant information, creating standardized but customizable letters, and mapping the locations of recipients. Provides a method for parsing and processing html code from online job postings of the American Political Science Association. (Available as a CRAN package)